Numerical Analysis On The Influence Of End Effect Of Geotechnical Specimen

The uniaxial compression test of rocks and triaxial test of soil are the most commonly used and the most basic tests in geotechnical tests. The mechanical properties of rocks and soil such as soil bearing capacity and constitutive relations can be obtained by measuring the stress, strain and other quantities. In both tests, due to the deformation of the machine indenter and the specimen is inconsistent, it will expect to friction constraint at the ends of specimen, and the specimen will inevitably be affected by end effects. So it’s important for processing the test data to study the problem of end effect.In actual tests, people have tried many ways to eliminate the influence of end effect. The effect is always less than ideal, while the desired end friction can be easily achieved by numerical simulation. So we can use numerical methods to discuss the influence of end effect. In this paper, the contact elements of ABAQUS software are used to simulate the contact action between the testing machine indenter and the ends of specimen. Stress distribution of points inside the specimen and the stress-strain relationship are simulated and calculated by setting different end condition such as free ends, completely restrained ends and ends with different frictional coefficients. By this way, the influence of end effect to test results is comprehensively discussed and recommendations are made to process the test data when the end effect is inevitable.Known from the analysis, the deformation of specimen in the two test simulations is uniform when the ends are free while the influence of end effect is significant in specimen with end constraint. The specimens are pressed drum-shaped at the end of tests, which are consistent with the actual test results. End effect will be a direct result of the uneven deformation of specimen, which makes the distribution of stress and strain inside the specimen complicate. The influence of end effect to the points inside the specimen decreases with increasing distance from the ends, which is the greatest at the ends of the specimen and the less on the middle region. The bearing capacity of the specimen, the reaction force-displacement curves and the equivalent plastic strain change in certain laws with the differences of ends frictions. For the marbles specimen of50mm×100mm, in the situation of completely restrain ends, the deformation of middle1/4region of the specimen is more uniform and largely unaffected by the impact of the end effect. So we can eliminate the impact of end effect by observing the middle1/4region of the specimen and at the same time, the force of this region is different from that applied to the sample by the test machine. So we need to reduce the force by11%, which is89%of the force applied by the machine. In triaxial tests, for the standard small triaxial clay specimen of39.1mm x80mm, the paper propose to eliminate the end effect by measuring the middle1/3region of the specimen, which is the same with the existing conclusions. But what is different is that stress of this1/3region is90%of the stress applied to the entire specimen. So we need to reduce10%of the stress applied to the specimen to ensure the experimental results to be reasonable and accurate.